An accountant in Vancouver just saved us $120K a year.
Prescott Thompson got tired of a close-management vendor. So he built his own using our internal AI platform Mesh.
Mesh is our internal AI productivity platform / AI coworker that gives every employee a highly configured agent connected to ZoomInfo’s internal context, systems, data, and workflows.
From accountants to sales reps to engineers - Mesh makes every employee a builder who can create AI tools tailored to their own workflows, rather than being a “button pusher” using tools made by someone else.
Prescott built PressClose - a full financial close management app with task management across 65 accountants, variance analysis, reconciliation tracking, audit trails, and direct SAP integration. The vendor we were paying $120K/year for was capped at 65 users, had no SAP integration, and our feature requests kept getting met with "it's on the roadmap."
Prescott's version does more AND it cost us $0 in incremental licensing.
Prescott isn't an outlier. He's one of HUNDREDS of people at ZoomInfo building micro-apps on top of ZI Mesh. A few that particularly impressed me:
- Zach Day, an AE in Chicago, is building Company Map - when he's traveling it shows every ICP-matching company within 5 miles, layered with intent signals. He's booking meetings between flights.
- Jessica Buchwald's marketing ops team built a campaign generator that took personalized sequence creation from 1.5 hours to 2 minutes per campaign.
- Sam D'Arcy in product built a voice-of-customer engine that reads every sales call and pulls out feature requests so PMs don't have to manually review hundreds of calls.
- Sarah Austin on our cash application team built an AR Payment Matcher that recovered a $16K payment that had been sitting unidentified for over a year.
None of these were on our roadmap. Nobody pinged me on slack asking for my permission. They just saw something broken in their day and fixed it.
We now have 550 micro-apps across the company, built on ZI Mesh.
All built BOTTOMS UP.
If you enable the person who understands the problem to also build the solution, EVERYTHING gets faster. You don’t waste time on spec docs, hand offs or have to translate the use case through 3 layers of PMs and engineers etc etc
For leaders implementing AI, our job is to make sure the data infrastructure and security exists AND the people who build something valuable get recognized for it.
Do those things and your team will surprise you.